RESEARCH ARTICLE
Re-conceptualizing stress: Shifting views on the consequences of stress and its effects on stress reactivity Jenny J. W. Liu, Kristin Vickers*☯, Maureen Reed☯, Marilyn Hadad☯ Department of Psychology, Ryerson University, Toronto, Ontario, Canada
a1111111111 a1111111111 a1111111111 a1111111111 a1111111111
☯ These authors contributed equally to this work. * [email protected]
Abstract Background
OPEN ACCESS Citation: Liu JJW, Vickers K, Reed M, Hadad M (2017) Re-conceptualizing stress: Shifting views on the consequences of stress and its effects on stress reactivity. PLoS ONE 12(3): e0173188. doi:10.1371/journal.pone.0173188 Editor: Alexandra Kavushansky, Technion Israel Institute of Technology, ISRAEL Received: October 13, 2016 Accepted: February 16, 2017 Published: March 8, 2017 Copyright: © 2017 Liu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: Data contains sensitive information that could be used to identify study participants, and is available upon request from the corresponding author of the study, Kristin Vickers ([email protected]). Funding: This research was supported by the Joseph-Armand Bombardier Canada Graduate Scholarships Program Master’s Scholarships from Social Sciences and Humanities Research Council of Canada ($17500 - http://www.sshrc-crsh.gc.ca/ funding-financement/programs-programmes/ fellowships/cgs_masters-besc_maitrise-eng.aspx).
The consequences of stress are typically regarded from a deficit-oriented approach, conceptualizing stress to be entirely negative in its outcomes. This approach is unbalanced, and may further hinder individuals from engaging in adaptive coping. In the current study, we explored whether negative views and beliefs regarding stress interacted with a stress framing manipulation (positive, neutral and negative) on measures of stress reactivity for both psychosocial and physiological stressors.
Method Ninety participants were randomized into one of three framing conditions that conceptualized the experience of stress in balanced, unbalanced-negative or unbalanced-positive ways. After watching a video on stress, participants underwent a psychosocial (Trier Social Stress Test), or a physiological (CO2 challenge) method of stress-induction. Subjective and objective markers of stress were assessed.
Results Most of the sampled population regarded stress as negative prior to framing. Further, subjective and objective reactivity were greater to the TSST compared to the CO2 challenge. Additionally, significant cubic trends were observed in the interactions of stress framing and stress-induction methodologies on heart rate and blood pressure. Balanced framing conditions in the TSST group had a significantly larger decrease in heart rate and diastolic blood pressure following stress compared to the positive and negative framing conditions.
Conclusion Findings confirmed a deficit-orientation of stress within the sampled population. In addition, results highlighted the relative efficacy of the TSST compared to CO2 as a method of stress provocation. Finally, individuals in framing conditions that posited stress outcomes in unbalanced manners responded to stressors less efficiently. This suggests that unbalanced
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
1 / 14
Re-conceptualizing stress
Competing interests: The authors have declared that no competing interests exist.
framing of stress may have set forth unrealistic expectations regarding stress that later hindered individuals from adaptive responses to stress. Potential benefits of alternative conceptualizations of stress on stress reactivity are discussed, and suggestions for future research are made.
Introduction The experience of stress has been known to diminish health, adjustment and well-being. Within the Transactional model of stress [1], stress is recognized as an interactive process between cognition and physiology and is defined in the current study as a stimulus, event, or situation that requires adaptation, and may challenge the existing equilibrium, whether psychologically or physically. As such, a stressful event can provoke a psychological and/or physiological response in various systems within an individual. Prolonged exposures to stress result in exhaustion of bodily resources, leaving one vulnerable to a myriad of mental and physical problems [2]. Thus, improving how individuals respond to stress has been the emphasis of much research. The detrimental effects of stress on health outcomes are well recognized in studies and in literature reviews [2, 3]. It is also widely assumed that stress is negative, unpredictable, and threatening [4, 5], although this assumption has not yet been tested empirically. However, researchers are increasingly noting that the experience of stress can engender both positive and negative outcomes over time [6]. For example, successful coping with everyday stress could serve as a mechanism to motivate achievement, personal growth, self-confidence and coping self-efficacy. Indeed, even unsuccessful attempts at coping could facilitate motivation and an opportunity to challenge one’s skills. Given that the consequences of stress can involve both positive and negative outcomes, actively (re)-framing what one perceives to be the outcomes of stressors may be key in facilitating adaptive responses to stress. Indeed, past studies have found that perceptions of the outcomes of stress are an important mediator in the relationship between stress and negative health outcomes [7]. Recent studies in stress [8–10] have successfully reduced stress responses by reframing the effects of anticipated stressors as more positive. However, the consequences of stress are hardly always positive. Consequently, a conceptualization of stress that is overly positive may not be beneficial in some contexts, such as when one’s views on stress are incongruent with reality [11]. For example, an overly positive message about stress might set forth expectations that one does not have to prepare for an upcoming event. This, coupled with the reality of experiencing an event that exerts demands on one’s resources, may then overwhelm the individual, hindering adaptive responses. Thus, it is important to examine how different expectations for the outcomes of stress may shape or affect individual reactivity to stress by first assessing a sampled population’s perceptions of stress, then manipulating these perceptions through different ways in which consequences of stress are framed. A considerable literature has been dedicated to the validation and comparison of varied methods of stress-induction. Recent advances in the comparative effects of multiple methods of stress-induction have highlighted the Trier Social Stress Test (TSST) as a robust and wellvalidated method of provoking psychosocial stress [12, 13]. Other methods of engendering stress have received less empirical attention in comparison. To date, only a handful of studies have used the carbon dioxide test (a single vital capacity inhalation of air enriched with 35% carbon dioxide) as a stress provocation [14]. These preliminary findings suggest that the
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
2 / 14
Re-conceptualizing stress
carbon dioxide (CO2) test may be a “very useful tool” ([15], p. 745) to provoke stress in the lab. Yet, no studies, to our knowledge, have compared the ability of the CO2 test, which can be thought of primarily as a physiological stressor, to engender the human stress response to that of a well-validated method of stress induction. Therefore, in the current research, we utilized both the CO2 test and the TSST, a reliable and robust method of stress-provocation. Our research aims were as follows: to examine individuals’ beliefs about stress; to study whether these beliefs interact with how individuals respond to stress; and to determine whether the stress response can be influenced by different ways in which stress outcomes are framed. As noted above, we also sought to compare stress reactivity to the CO2 test and the TSST. We hypothesized that the majority of the sampled population would conform to a negative perception of stress, and would believe that the consequences of stress are entirely negative. We also predicted that more positive beliefs about stress would facilitate reduced stressor reactivity via lower heart rate (HR), blood pressure (BP), and self-reported stress levels. We further hypothesized that by framing the effects of stress in a balanced manner to include both positive and negative information about stress outcomes, individuals would experience a faster recovery to baseline following stress, relative to framing conditions that described only the positive or the negative consequences of stress. Finally, in an exploratory fashion, we compared effects of the CO2 test on stress reactivity to that of the well-validated TSST psychosocial stressor.
Methods Participants A total of 97 participants successfully completed the study. Participants were enlisted through an Introductory Psychology Student Pool, which consists of a large cross-section of students from various disciplines, ethnicities, and years of study. Exclusion criteria were previous medical conditions such as asthma (that are contradicted with use of the CO2 test) and current levels of stress falling outside of two standard deviations of the population norm on the Perceived Stress Scale (PSS) [16]. Seven participants were removed following data cleaning; four participants were outliers, one participant’s physiological data was not available due to equipment malfunction, while two had missing data exceeding 20%. The final sample size for all subsequent analysis thus consisted of 90 participants, with a mean age of 21 (SD = 4.53). Out of the 90 participants, 27 (30%) were males and 63 (70%) were females. Thirty-three percent self-identified as Caucasian, 23% identified as East Asian, 30% identified as South Asian, while others identified as mixed or other race/ethnicity.
Materials and apparatus 1. Framing of stress. Participants received one of three conceptualizations of stress via a 5-minute framing video: balanced, unbalanced negative, or unbalanced positive. To maximize experimental control and to ensure that only the content and wording was manipulated across videos, all three videos were shot in the same setting, featured the same actor, and were of the same length. The actor was a Caucasian female graduate student dressed in professional, business-casual attire, and maintained the same neutral but professional attitude throughout all three videos. In the unbalanced-negative condition, the actor discussed the effects of stress by focusing on the negative effects of chronic stress, and advised viewers to engage in stress management to avoid these negative effects. The unbalanced-positive framing stress video focused selectively on the potential positive and adaptive effects of acute stress. Specifically, positive outcomes, such as increased motivation, vigilance, and efficient stress response times were
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
3 / 14
Re-conceptualizing stress
highlighted. Finally, in the balanced framing stress video, both the negative and positive consequences of acute stress were discussed. In addition, the video emphasized an individual’s choices in determining one’s response to stress, and posited stress management as a muscle that needs constant exercising in order to foster resiliency. 2. Methods of stress induction. Participants were randomized to undergo one of two stress-induction methods: the CO2 test [14] or the TSST [17]. The CO2 test has been extensively used in past research to as a potent panicogen in those with panic disorder. In healthy participants, the CO2 test has been advocated as a stress task [14] due to its induction of physiological changes along with mild subjective anxiety [14, 18]. The CO2 test involves a single vital-capacity breath of 35% CO2 and 65% oxygen that is held for four seconds prior to exhalation. The TSST is a widely used laboratory test to induce stress in both clinical and healthy participants [17]. The TSST consists of an anticipation/speech-preparation period, a speech delivery period in which participants speak on a topic, and a task period involving a backwards subtraction task, each task lasting ten minutes. To match the duration of the CO2 challenge, the TSST protocol was shortened to five minutes per task in the current study, similar to the modifications made in Skoluda et al.’s study [13].
Measures 1. Demographics information. Basic demographic information, such as gender, age, and ethnicity was collected at the beginning of the study. 2. Perception of stress. To gauge whether the sample indeed had a deficit-orientation of stress, a free association task was used in which participants were asked to name up to four words associated with ‘stress’. This task is a reliable and commonly used tool to index existing knowledge, norms, and perceptions on a given topic [19]. 3. Physiological stress response. Objective indications of stress responsivity consisted of heart rate (HR), systolic, and diastolic blood pressure (SBP; DBP) from the Biopac CNAP1 blood pressure system, a continuous, non-invasive measure of heart rate and blood pressure. Data were time-period marked for five time-points, including baseline one (t1; questionnaire period, approximately five to ten minutes in length), framing (t2; video period, five minutes in length), baseline two (t3; post-video questionnaire period, three minutes), stressor (t4; during either TSST or CO2 task period, both 15 minutes), and post-stressor (t5; recovery and questionnaire period, approximately five to ten minutes). 4. Subjective stress response. A Visual Analogue Scale (VAS) assessing for levels of stress was given to participants to rate their current levels of stress immediately following the stressinduction task. The scale asks participants to rank their current levels of stress, such that “0 means no stress or discomfort, where you do not feel disturbed at all, and 100 means extreme stress, or the worst kind of distress experienced imaginable”. The scale is a valid and reliable instrument for measuring characteristics and values along a continuum, and ranged from 0 (not at all stressed) to 100 (extremely stressed) [20].
Procedure Participants were invited for a 60-minute session at Ryerson University. They were randomly assigned to one of three framing conditions: (1) balanced framing condition (n = 30), (2) unbalanced-negative framing condition (n = 30), and (3) unbalanced-positive framing condition (n = 30). In addition, participants were also randomly assigned to one of two stress-induction conditions: (a) the Trier Social Stress Test (n = 45), and (b) the CO2 challenge (n = 45). Randomization resulted in fifteen participants assigned per group, for a total of six mutually-
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
4 / 14
Re-conceptualizing stress
exclusive experimental groups. Testing for participants occurred between 9am to 5pm on weekdays. Written informed consent was obtained from participants prior to the beginning of the study. This study was approved by Ryerson University’s Research Ethics Board (REB) in Toronto, Canada (REB 2014–188). Upon arrival, each participant was connected to the Biopac Continuous Non-invasive Arterial Pressure (CNAP)1 NIBP100D blood pressure system. His or her non-dominant arm was used for measures of blood pressure and heart rate via a finger clip transducer for pulse. Participants then answered a series of questions, and watched a 5-minute framing video on stress. Next, participants underwent one of two forms of stress-induction. Lastly, participants completed additional questions, including the VAS measure of subjective stress reactivity, and were disconnected from the CNAP machine, debriefed and compensated via course credit for their participation.
Statistical analysis All quantitative analyses were conducted using IBM’s Statistical Package for the Social Sciences (SPSS) Software1, version 21.0. The interactive effects of both experimental manipulations, including stress framing and stress-induction, were assessed via mixed-ANOVA, with specific emphasis on trend analysis. A post hoc power analysis using G Power for a mixed ANOVA (five by three by two interaction) design with an alpha of .05 was found to have a statistical power of 0.72 to detect an effect sizes of 0.25, considered a small to medium effect under Cohen’s guidelines [21].
Results Participant stress orientation and perception To gauge whether the sample indeed had a deficit-oriented approach to understanding the consequences of stress [5], two blind raters were asked to categorize each response from the free association task as positive (1), negative (-1), or neutral (0). The inter-rater reliability was 84.26% overall, and conflicts were resolved through discussion. Chi-square goodness-of-fit test determined whether negative, positive, and neutral orientations were equally rated amongst participants. Results revealed that stress orientation among participants were not equally distributed in the sampled population, X2 (2, N = 90) = 53.07, p < .001. The majority (66.67%) of participants rated the first word associated with stress as negative (see Table 1). Further, the overall ratings were also predominantly negative, with 76.47% of participants associating stress with more negative than positive words, while only 16.47% of participants considered stress to be more positive than negative.
Prior beliefs of stress on subjective reactivity to stress In order to determine if prior beliefs on stress may serve to confound later subjective responses to stressors, a one-way analysis of variance examined whether participants’ stress Table 1. Ratings of stress via word association task. Frequency
Percent
Cumulative Percent
Negative
60
66.7
66.7
Neutral
26
28.9
95.6
Positive
4
4.4
100.0
Table represents the proportions of the sampled population whose first word to describe stress was negative, neutral, or positive. doi:10.1371/journal.pone.0173188.t001
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
5 / 14
Re-conceptualizing stress
orientation affected outcomes on the VAS. Due to the large differences in group sizes between negative (n = 60), neutral (n = 26), and positive (n = 4), we pooled the neutral and positive into a single group. Tests of between-subjects effects did not reveal significant differences in subjective stress responses based on initial perceptions of stress: F(1, 88) = .311, p = .579, η2 = .004. Participants’ subjective responses to stress were similar despite initial perceptions of stress as negative (subjective stress response M = 41.02, SD = 27.88), or neutral/positive (M = 37.63, SD = 25.57).
Experimental manipulations on subjective stress reactivity A three-by-two analysis of variance with framing condition (Negative, n = 30, Balanced, n = 30; Positive, n = 30) and stress induction (TSST, n = 45; CO2, n = 45) as between-subject factors examined whether these experimental manipulations contributed to differences in subjective stress reactivity. Findings did not reveal a significant effect of framing condition on participants’ subjective response to stress: F(2, 84) = 1.287, p = .282, η2 = .03. Participants’ subjective responses to stress were similar across the negative framing (M = 40.83, SD = 29.71), balanced framing (M = 34.39, SD = 23.10), and positive framing (M = 44.79, SD = 27.89) conditions. However, results revealed a significant effect of stress condition on participants’ subjective response to stress: F(1, 84) = 29.29, p< .001, η2 = .26. Compared to those that underwent the CO2 stress induction (M = 26.44, SD = 22.87), those that underwent the TSST reported a significantly higher subjective stress response (M = 53.33, SD = 24.21) (see Fig 1). The interactive effects of stress framing and stress induction did not significantly affect subjective stress reactivity: F(2, 84) = .908, p = .407, η2 = .02.
Fig 1. Subjective stress responses across methods of stress induction. Figure illustrates the comparative efficacy of a well-validated method of stress induction and a novel method of physiological stress induction on self-reported ratings of stress. Error bars represent the standard error for each method of induction. doi:10.1371/journal.pone.0173188.g001
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
6 / 14
Re-conceptualizing stress
Experimental manipulations on objective measures of stress reactivity To examine differences in physiological stress parameters across experimental groupings, a five-by-three-by-two (5x3x2) mixed-measures ANOVA was used to examine the effects of time (from baseline to recovery) on different objective measures of stress (HR, SBP, DBP). The two between-subject factors for this ANOVA were framing condition (Negative, Balanced, Positive) and stress-induction condition (TSST, CO2). All physiological data were Log transformed (Log10) prior to analyses to normalize the data. Mauchly’s test indicated that the assumption of sphericity was violated for all three physiological stress parameters: heart rate (χ2(9) = 133.114, p
Re-conceptualizing stress: Shifting views on the consequences of stress and its effects on stress reactivity Jenny J. W. Liu, Kristin Vickers*☯, Maureen Reed☯, Marilyn Hadad☯ Department of Psychology, Ryerson University, Toronto, Ontario, Canada
a1111111111 a1111111111 a1111111111 a1111111111 a1111111111
☯ These authors contributed equally to this work. * [email protected]
Abstract Background
OPEN ACCESS Citation: Liu JJW, Vickers K, Reed M, Hadad M (2017) Re-conceptualizing stress: Shifting views on the consequences of stress and its effects on stress reactivity. PLoS ONE 12(3): e0173188. doi:10.1371/journal.pone.0173188 Editor: Alexandra Kavushansky, Technion Israel Institute of Technology, ISRAEL Received: October 13, 2016 Accepted: February 16, 2017 Published: March 8, 2017 Copyright: © 2017 Liu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: Data contains sensitive information that could be used to identify study participants, and is available upon request from the corresponding author of the study, Kristin Vickers ([email protected]). Funding: This research was supported by the Joseph-Armand Bombardier Canada Graduate Scholarships Program Master’s Scholarships from Social Sciences and Humanities Research Council of Canada ($17500 - http://www.sshrc-crsh.gc.ca/ funding-financement/programs-programmes/ fellowships/cgs_masters-besc_maitrise-eng.aspx).
The consequences of stress are typically regarded from a deficit-oriented approach, conceptualizing stress to be entirely negative in its outcomes. This approach is unbalanced, and may further hinder individuals from engaging in adaptive coping. In the current study, we explored whether negative views and beliefs regarding stress interacted with a stress framing manipulation (positive, neutral and negative) on measures of stress reactivity for both psychosocial and physiological stressors.
Method Ninety participants were randomized into one of three framing conditions that conceptualized the experience of stress in balanced, unbalanced-negative or unbalanced-positive ways. After watching a video on stress, participants underwent a psychosocial (Trier Social Stress Test), or a physiological (CO2 challenge) method of stress-induction. Subjective and objective markers of stress were assessed.
Results Most of the sampled population regarded stress as negative prior to framing. Further, subjective and objective reactivity were greater to the TSST compared to the CO2 challenge. Additionally, significant cubic trends were observed in the interactions of stress framing and stress-induction methodologies on heart rate and blood pressure. Balanced framing conditions in the TSST group had a significantly larger decrease in heart rate and diastolic blood pressure following stress compared to the positive and negative framing conditions.
Conclusion Findings confirmed a deficit-orientation of stress within the sampled population. In addition, results highlighted the relative efficacy of the TSST compared to CO2 as a method of stress provocation. Finally, individuals in framing conditions that posited stress outcomes in unbalanced manners responded to stressors less efficiently. This suggests that unbalanced
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
1 / 14
Re-conceptualizing stress
Competing interests: The authors have declared that no competing interests exist.
framing of stress may have set forth unrealistic expectations regarding stress that later hindered individuals from adaptive responses to stress. Potential benefits of alternative conceptualizations of stress on stress reactivity are discussed, and suggestions for future research are made.
Introduction The experience of stress has been known to diminish health, adjustment and well-being. Within the Transactional model of stress [1], stress is recognized as an interactive process between cognition and physiology and is defined in the current study as a stimulus, event, or situation that requires adaptation, and may challenge the existing equilibrium, whether psychologically or physically. As such, a stressful event can provoke a psychological and/or physiological response in various systems within an individual. Prolonged exposures to stress result in exhaustion of bodily resources, leaving one vulnerable to a myriad of mental and physical problems [2]. Thus, improving how individuals respond to stress has been the emphasis of much research. The detrimental effects of stress on health outcomes are well recognized in studies and in literature reviews [2, 3]. It is also widely assumed that stress is negative, unpredictable, and threatening [4, 5], although this assumption has not yet been tested empirically. However, researchers are increasingly noting that the experience of stress can engender both positive and negative outcomes over time [6]. For example, successful coping with everyday stress could serve as a mechanism to motivate achievement, personal growth, self-confidence and coping self-efficacy. Indeed, even unsuccessful attempts at coping could facilitate motivation and an opportunity to challenge one’s skills. Given that the consequences of stress can involve both positive and negative outcomes, actively (re)-framing what one perceives to be the outcomes of stressors may be key in facilitating adaptive responses to stress. Indeed, past studies have found that perceptions of the outcomes of stress are an important mediator in the relationship between stress and negative health outcomes [7]. Recent studies in stress [8–10] have successfully reduced stress responses by reframing the effects of anticipated stressors as more positive. However, the consequences of stress are hardly always positive. Consequently, a conceptualization of stress that is overly positive may not be beneficial in some contexts, such as when one’s views on stress are incongruent with reality [11]. For example, an overly positive message about stress might set forth expectations that one does not have to prepare for an upcoming event. This, coupled with the reality of experiencing an event that exerts demands on one’s resources, may then overwhelm the individual, hindering adaptive responses. Thus, it is important to examine how different expectations for the outcomes of stress may shape or affect individual reactivity to stress by first assessing a sampled population’s perceptions of stress, then manipulating these perceptions through different ways in which consequences of stress are framed. A considerable literature has been dedicated to the validation and comparison of varied methods of stress-induction. Recent advances in the comparative effects of multiple methods of stress-induction have highlighted the Trier Social Stress Test (TSST) as a robust and wellvalidated method of provoking psychosocial stress [12, 13]. Other methods of engendering stress have received less empirical attention in comparison. To date, only a handful of studies have used the carbon dioxide test (a single vital capacity inhalation of air enriched with 35% carbon dioxide) as a stress provocation [14]. These preliminary findings suggest that the
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
2 / 14
Re-conceptualizing stress
carbon dioxide (CO2) test may be a “very useful tool” ([15], p. 745) to provoke stress in the lab. Yet, no studies, to our knowledge, have compared the ability of the CO2 test, which can be thought of primarily as a physiological stressor, to engender the human stress response to that of a well-validated method of stress induction. Therefore, in the current research, we utilized both the CO2 test and the TSST, a reliable and robust method of stress-provocation. Our research aims were as follows: to examine individuals’ beliefs about stress; to study whether these beliefs interact with how individuals respond to stress; and to determine whether the stress response can be influenced by different ways in which stress outcomes are framed. As noted above, we also sought to compare stress reactivity to the CO2 test and the TSST. We hypothesized that the majority of the sampled population would conform to a negative perception of stress, and would believe that the consequences of stress are entirely negative. We also predicted that more positive beliefs about stress would facilitate reduced stressor reactivity via lower heart rate (HR), blood pressure (BP), and self-reported stress levels. We further hypothesized that by framing the effects of stress in a balanced manner to include both positive and negative information about stress outcomes, individuals would experience a faster recovery to baseline following stress, relative to framing conditions that described only the positive or the negative consequences of stress. Finally, in an exploratory fashion, we compared effects of the CO2 test on stress reactivity to that of the well-validated TSST psychosocial stressor.
Methods Participants A total of 97 participants successfully completed the study. Participants were enlisted through an Introductory Psychology Student Pool, which consists of a large cross-section of students from various disciplines, ethnicities, and years of study. Exclusion criteria were previous medical conditions such as asthma (that are contradicted with use of the CO2 test) and current levels of stress falling outside of two standard deviations of the population norm on the Perceived Stress Scale (PSS) [16]. Seven participants were removed following data cleaning; four participants were outliers, one participant’s physiological data was not available due to equipment malfunction, while two had missing data exceeding 20%. The final sample size for all subsequent analysis thus consisted of 90 participants, with a mean age of 21 (SD = 4.53). Out of the 90 participants, 27 (30%) were males and 63 (70%) were females. Thirty-three percent self-identified as Caucasian, 23% identified as East Asian, 30% identified as South Asian, while others identified as mixed or other race/ethnicity.
Materials and apparatus 1. Framing of stress. Participants received one of three conceptualizations of stress via a 5-minute framing video: balanced, unbalanced negative, or unbalanced positive. To maximize experimental control and to ensure that only the content and wording was manipulated across videos, all three videos were shot in the same setting, featured the same actor, and were of the same length. The actor was a Caucasian female graduate student dressed in professional, business-casual attire, and maintained the same neutral but professional attitude throughout all three videos. In the unbalanced-negative condition, the actor discussed the effects of stress by focusing on the negative effects of chronic stress, and advised viewers to engage in stress management to avoid these negative effects. The unbalanced-positive framing stress video focused selectively on the potential positive and adaptive effects of acute stress. Specifically, positive outcomes, such as increased motivation, vigilance, and efficient stress response times were
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
3 / 14
Re-conceptualizing stress
highlighted. Finally, in the balanced framing stress video, both the negative and positive consequences of acute stress were discussed. In addition, the video emphasized an individual’s choices in determining one’s response to stress, and posited stress management as a muscle that needs constant exercising in order to foster resiliency. 2. Methods of stress induction. Participants were randomized to undergo one of two stress-induction methods: the CO2 test [14] or the TSST [17]. The CO2 test has been extensively used in past research to as a potent panicogen in those with panic disorder. In healthy participants, the CO2 test has been advocated as a stress task [14] due to its induction of physiological changes along with mild subjective anxiety [14, 18]. The CO2 test involves a single vital-capacity breath of 35% CO2 and 65% oxygen that is held for four seconds prior to exhalation. The TSST is a widely used laboratory test to induce stress in both clinical and healthy participants [17]. The TSST consists of an anticipation/speech-preparation period, a speech delivery period in which participants speak on a topic, and a task period involving a backwards subtraction task, each task lasting ten minutes. To match the duration of the CO2 challenge, the TSST protocol was shortened to five minutes per task in the current study, similar to the modifications made in Skoluda et al.’s study [13].
Measures 1. Demographics information. Basic demographic information, such as gender, age, and ethnicity was collected at the beginning of the study. 2. Perception of stress. To gauge whether the sample indeed had a deficit-orientation of stress, a free association task was used in which participants were asked to name up to four words associated with ‘stress’. This task is a reliable and commonly used tool to index existing knowledge, norms, and perceptions on a given topic [19]. 3. Physiological stress response. Objective indications of stress responsivity consisted of heart rate (HR), systolic, and diastolic blood pressure (SBP; DBP) from the Biopac CNAP1 blood pressure system, a continuous, non-invasive measure of heart rate and blood pressure. Data were time-period marked for five time-points, including baseline one (t1; questionnaire period, approximately five to ten minutes in length), framing (t2; video period, five minutes in length), baseline two (t3; post-video questionnaire period, three minutes), stressor (t4; during either TSST or CO2 task period, both 15 minutes), and post-stressor (t5; recovery and questionnaire period, approximately five to ten minutes). 4. Subjective stress response. A Visual Analogue Scale (VAS) assessing for levels of stress was given to participants to rate their current levels of stress immediately following the stressinduction task. The scale asks participants to rank their current levels of stress, such that “0 means no stress or discomfort, where you do not feel disturbed at all, and 100 means extreme stress, or the worst kind of distress experienced imaginable”. The scale is a valid and reliable instrument for measuring characteristics and values along a continuum, and ranged from 0 (not at all stressed) to 100 (extremely stressed) [20].
Procedure Participants were invited for a 60-minute session at Ryerson University. They were randomly assigned to one of three framing conditions: (1) balanced framing condition (n = 30), (2) unbalanced-negative framing condition (n = 30), and (3) unbalanced-positive framing condition (n = 30). In addition, participants were also randomly assigned to one of two stress-induction conditions: (a) the Trier Social Stress Test (n = 45), and (b) the CO2 challenge (n = 45). Randomization resulted in fifteen participants assigned per group, for a total of six mutually-
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
4 / 14
Re-conceptualizing stress
exclusive experimental groups. Testing for participants occurred between 9am to 5pm on weekdays. Written informed consent was obtained from participants prior to the beginning of the study. This study was approved by Ryerson University’s Research Ethics Board (REB) in Toronto, Canada (REB 2014–188). Upon arrival, each participant was connected to the Biopac Continuous Non-invasive Arterial Pressure (CNAP)1 NIBP100D blood pressure system. His or her non-dominant arm was used for measures of blood pressure and heart rate via a finger clip transducer for pulse. Participants then answered a series of questions, and watched a 5-minute framing video on stress. Next, participants underwent one of two forms of stress-induction. Lastly, participants completed additional questions, including the VAS measure of subjective stress reactivity, and were disconnected from the CNAP machine, debriefed and compensated via course credit for their participation.
Statistical analysis All quantitative analyses were conducted using IBM’s Statistical Package for the Social Sciences (SPSS) Software1, version 21.0. The interactive effects of both experimental manipulations, including stress framing and stress-induction, were assessed via mixed-ANOVA, with specific emphasis on trend analysis. A post hoc power analysis using G Power for a mixed ANOVA (five by three by two interaction) design with an alpha of .05 was found to have a statistical power of 0.72 to detect an effect sizes of 0.25, considered a small to medium effect under Cohen’s guidelines [21].
Results Participant stress orientation and perception To gauge whether the sample indeed had a deficit-oriented approach to understanding the consequences of stress [5], two blind raters were asked to categorize each response from the free association task as positive (1), negative (-1), or neutral (0). The inter-rater reliability was 84.26% overall, and conflicts were resolved through discussion. Chi-square goodness-of-fit test determined whether negative, positive, and neutral orientations were equally rated amongst participants. Results revealed that stress orientation among participants were not equally distributed in the sampled population, X2 (2, N = 90) = 53.07, p < .001. The majority (66.67%) of participants rated the first word associated with stress as negative (see Table 1). Further, the overall ratings were also predominantly negative, with 76.47% of participants associating stress with more negative than positive words, while only 16.47% of participants considered stress to be more positive than negative.
Prior beliefs of stress on subjective reactivity to stress In order to determine if prior beliefs on stress may serve to confound later subjective responses to stressors, a one-way analysis of variance examined whether participants’ stress Table 1. Ratings of stress via word association task. Frequency
Percent
Cumulative Percent
Negative
60
66.7
66.7
Neutral
26
28.9
95.6
Positive
4
4.4
100.0
Table represents the proportions of the sampled population whose first word to describe stress was negative, neutral, or positive. doi:10.1371/journal.pone.0173188.t001
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
5 / 14
Re-conceptualizing stress
orientation affected outcomes on the VAS. Due to the large differences in group sizes between negative (n = 60), neutral (n = 26), and positive (n = 4), we pooled the neutral and positive into a single group. Tests of between-subjects effects did not reveal significant differences in subjective stress responses based on initial perceptions of stress: F(1, 88) = .311, p = .579, η2 = .004. Participants’ subjective responses to stress were similar despite initial perceptions of stress as negative (subjective stress response M = 41.02, SD = 27.88), or neutral/positive (M = 37.63, SD = 25.57).
Experimental manipulations on subjective stress reactivity A three-by-two analysis of variance with framing condition (Negative, n = 30, Balanced, n = 30; Positive, n = 30) and stress induction (TSST, n = 45; CO2, n = 45) as between-subject factors examined whether these experimental manipulations contributed to differences in subjective stress reactivity. Findings did not reveal a significant effect of framing condition on participants’ subjective response to stress: F(2, 84) = 1.287, p = .282, η2 = .03. Participants’ subjective responses to stress were similar across the negative framing (M = 40.83, SD = 29.71), balanced framing (M = 34.39, SD = 23.10), and positive framing (M = 44.79, SD = 27.89) conditions. However, results revealed a significant effect of stress condition on participants’ subjective response to stress: F(1, 84) = 29.29, p< .001, η2 = .26. Compared to those that underwent the CO2 stress induction (M = 26.44, SD = 22.87), those that underwent the TSST reported a significantly higher subjective stress response (M = 53.33, SD = 24.21) (see Fig 1). The interactive effects of stress framing and stress induction did not significantly affect subjective stress reactivity: F(2, 84) = .908, p = .407, η2 = .02.
Fig 1. Subjective stress responses across methods of stress induction. Figure illustrates the comparative efficacy of a well-validated method of stress induction and a novel method of physiological stress induction on self-reported ratings of stress. Error bars represent the standard error for each method of induction. doi:10.1371/journal.pone.0173188.g001
PLOS ONE | DOI:10.1371/journal.pone.0173188 March 8, 2017
6 / 14
Re-conceptualizing stress
Experimental manipulations on objective measures of stress reactivity To examine differences in physiological stress parameters across experimental groupings, a five-by-three-by-two (5x3x2) mixed-measures ANOVA was used to examine the effects of time (from baseline to recovery) on different objective measures of stress (HR, SBP, DBP). The two between-subject factors for this ANOVA were framing condition (Negative, Balanced, Positive) and stress-induction condition (TSST, CO2). All physiological data were Log transformed (Log10) prior to analyses to normalize the data. Mauchly’s test indicated that the assumption of sphericity was violated for all three physiological stress parameters: heart rate (χ2(9) = 133.114, p
